The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/rename_project.py uses the subprocess.run function to execute Git commands and the GitHub CLI (gh). The tool implements security controls by validating the target project name against a restrictive regular expression (^[A-Za-z0-9][A-Za-z0-9._-]*$), which effectively prevents shell command injection.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external project files, which constitutes a potential attack surface for indirect prompt injection.
Ingestion points: Files are read from the local project root in scripts/rename_project.py to generate a rename plan.
Boundary markers: The skill does not currently use specific delimiters to wrap content read from files when presenting the plan to the agent.
Capability inventory: The skill possesses file-write capabilities and the ability to execute Git and GitHub CLI commands through subprocesses.
Sanitization: While command arguments are validated, the textual content read from project files is not sanitized, as the tool's purpose is literal string replacement.
[DATA_EXPOSURE]: The utility scripts/profile_store.py manages persistent user configuration in standard local directories (e.g., ~/.config/agent-skills). It includes a security filter that explicitly rejects attempts to store sensitive keys such as tokens, secrets, or passwords within the profile records.

lov-rename-project